Summary
GPT Researcher is a multi-agent research framework that parallelizes web and local document retrieval to generate long, sourced research reports suitable for academic and industry investigations.
Features
- Planner/Executor/Publisher agent architecture enabling deep, tree-like exploration (Deep Research).
- Integrated web scraping, image sourcing and context memory, with export options for PDF/Word/Markdown.
- Includes TaskBench, MCP integration and multiple deployment options (PIP, Docker, NextJS frontends).
Use Cases
- Automated market or academic research that aggregates evidence from many sources into auditable reports.
- Domain-specific research over local documents and corpora.
- Benchmarking LLM-driven automation workflows with TaskBench.
Technical Details
- Implemented primarily in Python; frontend available in lightweight static and production NextJS variants.
- Supports
local/huggingface/hybridretrieval and inference modes and an MCP client extension. - Licensed under Apache-2.0 with active community contributions and comprehensive docs.